FOLIAR SCLEREIDS IN DIONYSIA (PRIMULACEAE) FROM A PHYLOGENETIC PERSPECTIVE

2006·0.84 MB·

by  Trift, I

EDINBURGH JOURNAL OF BOTANY 63 (1): 21–48 (2006) 21
doi:10.1017/S096042860600045X © Trustees of the Royal Botanic Garden Edinburgh (2006)
Issued 9 August 2006
FOLIAR SCLEREIDS IN DIONYSIA (PRIMULACEAE)
FROM A PHYLOGENETIC PERSPECTIVE
I. TRIFT & A. A. ANDERBERG1
Sclereid presence and distribution were studied in leaves from 50 species of Dionysia
(Primulaceae). Of 16 species not previously investigated, 11 were shown to possess
sclereids. The sclereids are dermal, terminal or diffused (according to the position in
the leaf) and develop from different tissues, and hence are not homologous structures.
The presence of different types of sclereids in different species is discussed. Sclereid
characters are optimized on a cladogram based on an analysis of three DNA regions.
Some clades are associated with certain types of sclereids. Terminal sclereids are most
common in Iran, and diffused sclereids in Afghanistan. The evolution of sclereids in
the genus is discussed, and a redefinition of what can be called a sclereid in Dionysia is
presented.
Keywords. Dionysia, leaf sclereids, phylogeny, Primula, Primulaceae, stone cells.
INTRODUCTION
The genus Dionysia Fenzl comprises 51 species, of which 13 have been discovered
since 1976. This genus of tuft- or cushion-forming xerophytic subshrubs has been
shown to have evolved within the larger genus Primula L., its sister group being
Primula subgen. Sphondylia (Mast et al., 2001; Trift et al., 2002, 2004). A phylo-
genetic study of Dionysia based on three DNA regions, morphology and biogeo-
graphy has recently been undertaken (Trift et al., 2004). The resulting cladograms
indicated that the sections and subsections of Dionysia used in previous divisions of
the genus (Wendelbo, 1961, 1964) are not monophyletic. Trift et al. (2004) found
that many character states were size dependent. More specifically, characters like
scape and petiole were reduced or absent in small species. The need for more size-
independent morphological characters was noted in the conclusion. The results of
Bokhari & Wendelbo (1976, 1985) and Bokhari (1988) showed that the presence of
leaf sclereids was one such potential character.
Leaf sclereids in angiosperms can be described according to their position in the
leaf, their ontogeny, or their shape (Foster, 1949). The many forms of sclereid found
in plants have induced a rich and varied terminology. This study attempts to follow
the terminology of Dickison (2000), which is representative of those adopted by
most authors. Sclereids are referred to as being either dermal, terminal or diffused
depending on their position in the leaf. Dermal sclereids are found in the epidermis;
Department of Phanerogamic Botany, Swedish Museum of Natural History, PO Box 50007, 104 05
Stockholm, Sweden. 1Author for correspondence; e-mail: [email protected]
22 I. TRIFT & A. A. ANDERBERG
terminal sclereids are found at the ends of the veins, and diffused sclereids are found
dispersed in the mesophyll. The origins of sclereids are related to their position in
mature leaves. Sclereids with terminal positions normally originate from the trache-
ary elements, but they can in some cases originate from the parenchyma and become
positioned next to the vein ends during leaf ontogeny (Rao, 1951). This only applies
to some terminal sclereids; there is no indication that dermal sclereids can develop
from anything but epidermal cells, and diffused sclereids from anything but paren-
chymatic cells (Rao, 1951). When sclereids develop from different tissues, they
cannot be considered homologous. However, they form in response to a common
factor, hormonal or otherwise regulatory, that directs the development of the differ-
entiating cells in all plant tissues. Bokhari & Wendelbo (1976, 1985) and Bokhari
(1988) found dermal, terminal and diffused sclereids in different species, and it
is interesting to see that a relatively small genus such as Dionysia possesses all
three kinds. In Bokhari & Wendelbo (1985) the xeromorphic features in Dionysia
were described in detail, and they noted that sclerenchyma in leaves is thought
to protect against damage from wilting. They made the following observations on
sclerenchyma:
I. Dermal sclereids in Dionysia were strongly sclerified epidermal cells, long and
narrow as opposed to the normal barrel-shaped, thin-walled cells. They were always
of the macrosclereid type when described according to shape. Dermal sclereids were
found only on the exposed side of the leaves.
II. Terminal sclerenchyma in Dionysia were of three types: polymorphic sclereids,
fibrous caps, and tracheoids.
a. Terminal polymorphic sclereids were not confined to the ends of the veins, but
also found along the veins. These sclereids varied considerably in size, form and
degree of branching, and could accompany the entire vascular tissue.
b. Fibrous caps (consisting of groups of fibres) were located on the adaxial side of
the veins in the upper half of the leaves.
c. Terminal tracheoids were described as idioblasts, resembling tracheids in their
pitted walls, but differing from typical tracheary elements in their form (shorter
and wider), size (larger) and position. Tracheoids are assumed to have different
functions from typical tracheary elements and in other genera they are typically
not found in association with veins (Bokhari & Burtt, 1970). There were signifi-
cant differences between the tracheoids described from Cyrtandra (Gesneriaceae)
(Bokhari & Burtt, 1970) and those later described from Dionysia (Bokhari &
Wendelbo, 1976, 1985; Bokhari, 1988).
III. Diffused sclereids in Dionysia were found dispersed singly or in groups in the leaf
mesophyll. Six types of diffused sclereid were described in Dionysia. Some of these
types occurred together in the same leaf, but none of them was very common. Each
type was found only in one to three species.
FOLIAR SCLEREIDS IN DIONYSIA 23
a. Brachysclereids: short isodiametric cells resembling parenchyma cells.
b. Fusiform sclereids: elongated and pointed at the ends.
c. Filiform sclereids: very long and narrow, showing very intrusive growth.
d. Macrosclereids: somewhat elongated cells, roughly cylindrical, with uneven
secondary walls.
e. Polymorphic sclereids: bent, curved or branched, of all sizes.
f. Vesiculose sclereids: little different in shape from neighbouring parenchyma cells
and not showing intrusive growth.
Of the 51 Dionysia species, 16 were not included by Bokhari & Wendelbo (1976,
1985) or Bokhari (1988), either due to lack of material (D. hissarica, D. involucrata,
D. kossinskyi), or because they were not described at that time (D. aubrietioides, D.
bazoftica, D. iranica, D. khatamii, D. khuzistanica, D. lurorum, D. mozaffarianii, D.
sarvestanica, D. sp. nov. 1, D. sp. nov. 2, D. sp. nov. 3, D. sp. nov. 4, D. sp. nov. 5).
When the sclereid character states from Bokhari & Wendelbo (1976) and Bokhari
(1988) were plotted on the cladogram based on three DNA regions (Trift et al.,
2004), terminal and dermal leaf sclereids (not terminal tracheoids) appeared to have
evolved late in the history of the genus and only in western species from Iran and
Iraq. Diffused leaf sclereids appeared only in species growing in Afghanistan and
Pakistan. Species with peripheral distributions, growing in Oman, Tadzjikistan,
Turkmenistan and Turkey, either lacked sclereids or had not been investigated.
Thus, the absence or presence of leaf sclereids was a potentially informative charac-
ter that could merit another investigation. A new study with new samples could also
increase the knowledge on infraspecific variation.
The aim of this study was to investigate the presence of leaf sclereids in Dionysia,
particularly in species that were not available to Bokhari and Wendelbo, and to
compare the distribution of each type with recent DNA-based hypotheses of the
phylogeny of the genus.
MATERIAL
Dry leaves (herbarium material or silica dried) were used. Dickison (2000) noted
that sclereids originate late in the ontogeny of the leaf, and fully developed sclereids
are found only in mature leaves. The best way to make sure that all sclereids were
fully developed was to use the wilted, remaining leaves from previous seasons that
are found on the older parts of most Dionysia species. Only one species was not
included in the present study, due to lack of material (D. lacei). With four excep-
tions, only one specimen from each species was examined. The voucher numbers are
given in Table 1, which also shows the tentative division of the genus from Trift
et al. (2004).
METHODS
In their study Bokhari & Wendelbo (1976) referred to the earlier study by Bokhari &
Burtt (1970) for a description of the methods, which in turn referred to Bokhari’s
24 I. TRIFT & A. A. ANDERBERG
TABLE 1. Dionysia species, in the order of the tentative new classification of the genus (Trift
et al., 2004), with voucher numbers for sampled specimens
Species Voucher number, comments
CLADE A
D. balsamea Wendelbo & Rech.f. No 21767, 919111 (E)
D. hissarica Lipsky JJH 918037, Halda 92, 960605, Gothenburg
Bot. g. cult.
CLADE B
Southern and eastern species, clade C
D. mira Wendelbo 2002-02-01, leg. Reginald Victor (S)
D. denticulata Wendelbo 1999-1320f, Gothenburg Bot. g. cult.
D. tapetodes Bunge MK 8809/1, Gothenburg Bot. g. cult. (no
sclereids) and Hewer 1164, Gothenburg
Bot. g. cult. (fili, fusi) and T4Z 1086-1,
Gothenburg Bot. g. cult. (brac, macr,
fili, fusi)
D. kossinskyi Czerniak. Gaudan 29.04.1912, V. Lipsky (LE)
D. lindbergii Wendelbo Grey-Wilson & Hewer 1304, 30.06.1971 (GB)
D. microphylla Wendelbo Trift 57, 1999-04-19 (S)
D. involucrata Zapr. Trift 201 (S)
D. hedgei Wendelbo Grey-Wilson & Hewer 836, 12.05.1971 (GB)
D. viscidula Wendelbo Grey-Wilson & Hewer 1305, Gothenburg
Bot. g. cult.
D. afghanica Grey-Wilson 2003-893, Gothenburg Bot. g. cult.
and Kammerlander Czdenek Zvolaner
MK 2003-884, Gothenburg Bot. g. cult.
Hybrid individual
D. freitagii Wendelbo Grey-Wilson & Hewer 8479, Gothenburg
Bot. g. cult.
Southern and eastern species, part clade E
D. paradoxa Wendelbo Wendelbo 744, Gothenburg Bot. g. cult.
(no sclereids) and Grey-Wilson & Hewer
1026 (E) (macr, poly) (photo)
D. lacei Clay Not examined
D. saponacea Wendelbo & Rech.f. Rechinger 19092 (S)
Western species, clade F
D. archibaldii Wendelbo Archibald 3010, 05.08.1966 (GB)
D. esfandiarii Wendelbo SLIZE 259:4, Gothenburg Bot. g. cult.
D. sp. nov. 1 T4Z 35 (UPS)
D. revoluta subsp. revoluta Boiss. SLIZE 252, Gothenburg Bot. g. cult.
D. oreodoxa Bornm. Zschummel 300-6 (UPS)
D. rhaptodes Bunge Wendelbo & Foroughi 15880, 12.04.1975 (GB)
D. teucrioides Davis & Wendelbo Trift 200 (S)
D. bornmülleri (Pax) Clay Davis 42833, 1966-05-11, 919118 (E)
D. leucotricha Bornm. SLIZE 289, Gothenburg Bot. g. cult.
D. aretioides (Lehm.) Boiss. Trift 56, 1999-04-19 (S)
and SLIZE 35, Gothenburg Bot. g. cult.
FOLIAR SCLEREIDS IN DIONYSIA 25
TABLE 1. (Cont’d).
Species Voucher number, comments
D. khatamii Mozaff. T4Z 13 (UPS)
D. janthina Bornm. & Wink. 19932488 (E)
D. curviflora Bunge SLIZE 42:268 (UPS)
Western species, clade G
D. lurorum Wendelbo Zschummel 00-40-3, Gothenburg Bot. g. cult.
D. aubrietioides Jamzad & Mozaff. T4Z 133 (UPS)
D. iransharii Wendelbo K.H. Rechinger 47433 (S)
D. iranica Jamzad T4Z 119 (UPS)
D. khuzistanica Jamzad SLIZE 176:4, Gothenburg Bot. g. cult.
D. zagrica Grey-Wilson Hewer 2023 (GB)
D. caespitosa (Duby) Boiss. Zschummel 300-14 (UPS)
D. mozaffarianii Lidén SLIZE 232:8, Gothenburg Bot. g. cult.
D. odora Fenzl 99-0713-1, Gothenburg Bot. g. cult.
D. gaube Bornm. Archibald 1633 (GB)
D. haussknechtii Bornm. & Strauss SLIZE 322:2, Gothenburg Bot. g. cult.
D. lamingtonii Stapf Hewer H1909, 13.04.1973 (GB)
D. sp. nov. 2 Zschummel 300-18 (UPS)
D. sawyeri (Watt) Wendelbo Alexeenko 2722 (LE)
(D. bachtiarica cotype Bornm. et al. ex.)
D. termeana Wendelbo Hewer H1987, 08.05.1973 (UPS)
D. bryoides Boiss. SLIZE 236:2, Gothenburg Bot. g. cult.
D. diapensiifolia Boiss. SLIZE 253:2, Gothenburg Bot. g. cult. (photo)
and Wendelbo & Fouroghi 17548, 30.05.1975
(TARI)
D. sarvestanica Jamzad & Grey-Wilson SLIZE 241:3, Gothenburg Bot. g. cult.
D. michauxii (Duby) Boiss. SLIZE 254:3, Gothenburg Bot. g. cult.
D. sp. nov. 3 T4Z 125, isotype (UPS)
D. sp. nov. 4 Wendelbo & Assadi 16761, holotype (TARI)
D. sp. nov. 5 Dehgani 5466, holotype (TARI)
Primula subgen. Sphondylia
P. verticillata Forssk. Trift 202, 04.09.2002, cult. (S)
P. simensis Hochst. St Andrews 1997-0923, Gothenburg Bot. g. cult.
P. floribunda Wall. Trift 1999-04-19 (S)
study on Limonium leaves (Bokhari, 1970). In our study we made no leaf macera-
tions or transections. The method for leaf clearing and staining was slightly modified
from that of Bokhari (1970). Leaves were placed in glass dishes with 10% KOH
solution overnight. The KOH solution was then replaced with water that was
exchanged for fresh water three times in 1 hour. The water was then replaced with
4.4% sodium hypochlorite (a common household bleach) and the leaves soaked until
they were transparent. The bleaching time varied with the size and texture of the
leaves from 20 minutes to 5 hours. Once the leaves were bleached, they were boiled
in water for 3 minutes. This process loosened the epidermis. The boiled leaves were
26 I. TRIFT & A. A. ANDERBERG
placed in a drop of water on a microscope slide and the epidermis was peeled off.
The peeled leaves were stained with safranin in water. After 5 minutes, excess safra-
nin was removed with tissue paper, and the leaves were washed with water and then
with 95% ethanol. This was done by adding the washing liquid with a pipette to the
slide surface, and then carefully removing it with tissue paper. The leaves were then
dehydrated with 99.5% ethanol in the same manner. To make a permanent slide,
a few drops of euparal were placed on the peeled, stained leaves and topped with a
cover slide, the edges of which were sealed with varnish. The prepared slides were
photographed with a digital camera mounted on a light microscope.
When removing the epidermis care was taken not to tear off the edges of the leaf
mesophyll, which adhere more strongly to the epidermis than do the flat sides. This
was especially important for leaves with strongly revolute leaf margins. Large, soft
leaves were the most difficult to prepare and could fall apart completely if not
treated gently. The presence of large numbers of sclereids generally made the leaves
easy to prepare as they stabilized the tissue. The powdery farina present on the leaf
surface in some species can be confused with sclereids as it loosens during prepara-
tion, but this problem is easily avoided if the farina is examined separately before the
leaves are prepared. When measuring sclereid bundles the maximum width was
taken at the mid vein, and the smaller measurement on the largest side vein towards
the tip of the leaf. There were always much smaller bundles on veinlets that were not
measured.
Sclereid characters were plotted onto a cladogram from Trift et al. (2004) using
MacClade software (Maddison & Maddison, 1992) (Fig. 1). This cladogram was the
result of a Bayesian inference analysis of the combined sequence data from three
DNA regions: rps16, trnL–trnF and ITS. The cladogram was based on an analysis
of sequences from 39 Dionysia species; hence 12 species were not included. These
were assigned hypothetical positions based on biogeography and their affinities
according to previous classifications based on morphology (Grey-Wilson, 1989).
These have been drawn with dotted lines to indicate that their positions are uncer-
tain. The support values for the tree are not included here. Some branches had a
support lower than 67%; these are drawn with zigzag lines. All other branches had a
support of 74–100%. The sclereid results were also plotted on a distribution map
from Trift et al. (2004) (Fig. 2).
RESULTS
Dermal (I), terminal (II) and diffused sclereids (III) were found in Dionysia, and
no sclereids were found in Primula subgen. Sphondylia. The category ‘terminal
tracheoids’ was not used; instead ‘tracheoid’ was used as an adjective to describe
vein ends resembling those in leaves of D. lindbergii. The vein ends were described
as plain, thick or tracheoid. We will return to this matter in the discussion. We
classified the diffused sclereids into five categories:
FOLIAR SCLEREIDS IN DIONYSIA 27
a. brachysclereids
b. fusiform sclereids
c. filiform sclereids
d. macrosclereids
e. polymorphic sclereids.
FIG. 1. Sclereid characters plotted on a cladogram taken from Trift et al. (2004). Species
in parentheses were not included in this study. Sclereid characters in parentheses were not
encountered in this study. Zigzag lines indicate clades with a weak support (67% or less in the
Bayesian inference analysis). Twelve species were not included in Trift et al. (2004) and their
positions in the tree are not known. They have been positioned with dotted lines based on
their affinities as suggested by their respective authors when first described.
28 I. TRIFT & A. A. ANDERBERG
4).
0
0
2
(
al.
t
e
rift
T
m
o
r
f
n
e
k
a
t
p
a
m
n
o
uti
b
ri
st
di
a
n
o
d
e
ott
pl
s
r
e
ct
a
r
a
h
c
d
ei
r
e
cl
S
2.
.
G
I
F
FOLIAR SCLEREIDS IN DIONYSIA 29
Unlike Bokhari & Wendelbo (1976) and Bokhari (1988), we did not include the
category vesiculose sclereids as these were reported only from D. lacei, of which we
had no material.
I. Dermal sclereids (Figs 3, 4) were found in D. curviflora and D. microphylla. In
both species, sclereids were confined to the distal part of the lower side of the leaves
(the exposed part). Dionysia curviflora had elongated, thick-walled dermal sclereids,
whereas the dermal sclereids in D. microphylla were rhomboid, of the same shape as
surrounding epidermal cells.
II. Terminal sclereids, found at the ends of veins and along the veins, were found in
many species. These sclereids resemble the vessel elements in their initial annular
thickening, but are more irregular in shape and have thicker walls. Table 2 lists
measurements of bundle width in relation to the width of the leaves. By definition, a
thick-walled sclerenchyma cell that is long and has tapered ends and is associated
with the vessels is a fibre (Dickison, 2000). Some of the terminal sclereids found are
therefore more correctly called fibres. Bokhari & Wendelbo (1985) called this feature
fibrous caps.
Species with leaves containing a few terminal sclereids along the veins were D.
bryoides, D. hissarica, D. revoluta subsp. revoluta, D. saponacea and D. sp. nov. 1.
FIG. 3. Elongate dermal sclereids in Dionysia curviflora. Pale dermal cells are not sclerified.
See Table 1 for accession details for Figs 3–24.
30 I. TRIFT & A. A. ANDERBERG
FIG. 4. Dermal sclereids in Dionysia microphylla. Note that the epidermal cells above the leaf
tip have thicker walls than the cells to the left (arrows).
TABLE 2. Measurements on the width of bundles of terminal sclereids, in order by increasing
maximum relative width
Relative width of
Species Bundle width (mm) Leaf width (mm) bundle/leaf (%)
D. caespitosa 0.02–0.05 2.0–2.8 0.7–2.3
D. lurorum 0.04–0.06 2.0–3.0 1.3–3.0
D. sp. nov. 5 0.04–0.07 1.6–2.0 2.2–3.9
D. aubrietioides 0.05–0.08 2.0–2.6 1.9–4.0
D. diapensiifolia 0.04–0.09 2.0–2.2 2.0–4.1
D. odora 0.06–0.09 2.0–2.2 2.7–4.1
D. iranica 0.03–0.06 1.2–1.5 2.0–5.0
D. gaube 0.03–0.05 1.0–1.4 2.1–5.0
D. michauxii 0.03–0.06 1.1–1.6 2.1–5.5
D. sp. nov. 3 0.02–0.06 0.7–1.8 2.2–5.7
D. sarvestanica 0.04–0.10 0.9–1.2 3.6–9.1
D. zagrica 0.04–0.10 0.9–1.4 3.6–10.0
D. sp. nov. 4 0.04–0.10 0.7–1.0 4.0–10.0
D. khuzistanica 0.07–0.19 0.8–1.5 6.6–13.8
D. lamingtonii 0.06–0.11 0.5–1.3 4.6–22.0